BACKGROUND: Area-level socio-economic factors are significantly related to a population's health. This study investigates how school district-level factors affect the initiation of alcohol drinking of Danish adolescents. METHODS: A survey sample of 11,223 female and male pupils in the 7th grade from 447 schools across Denmark was analysed for the outcome variable drinking initiation and a number of individual level predictors. Aggregated variables on school district level were created from national registry data for education, occupational level and household savings of residents, type of housing and land use characteristics. RESULTS: About 40% of all respondents (45.8% males and 35.2% females) had ever drunk more than one glass of alcoholic beverage. Mixed-effects logistic regression showed that significant individual level predictors for drinking initiation were male gender, a lower performance at school, perceived peer group drinking and the perceived daily drinking of the father. On school district level, adolescents were more likely to initiate alcohol consumption in school districts with higher farming land use and less likely in those with higher proportion of private apartment buildings. Other school district factors were not associated with drinking initiation when controlled for individual level factors. CONCLUSIONS: The impact of socio-economic variables at school district level seems to be smaller in the welfare state of Denmark than known for other countries. However, residence in rural areas may be a direct disadvantage for youth, indicating a need for region-specific prevention programmes.
BACKGROUND: Area-level socio-economic factors are significantly related to a population's health. This study investigates how school district-level factors affect the initiation of alcohol drinking of Danish adolescents. METHODS: A survey sample of 11,223 female and male pupils in the 7th grade from 447 schools across Denmark was analysed for the outcome variable drinking initiation and a number of individual level predictors. Aggregated variables on school district level were created from national registry data for education, occupational level and household savings of residents, type of housing and land use characteristics. RESULTS: About 40% of all respondents (45.8% males and 35.2% females) had ever drunk more than one glass of alcoholic beverage. Mixed-effects logistic regression showed that significant individual level predictors for drinking initiation were male gender, a lower performance at school, perceived peer group drinking and the perceived daily drinking of the father. On school district level, adolescents were more likely to initiate alcohol consumption in school districts with higher farming land use and less likely in those with higher proportion of private apartment buildings. Other school district factors were not associated with drinking initiation when controlled for individual level factors. CONCLUSIONS: The impact of socio-economic variables at school district level seems to be smaller in the welfare state of Denmark than known for other countries. However, residence in rural areas may be a direct disadvantage for youth, indicating a need for region-specific prevention programmes.